Conventionally, indoor mapping is relegated to manually targeting specific Wi-Fi signals, Bluetooth beacons, or visual fiducial (e.g. barcode) markers. These techniques usually need significant initial human annotation during setup and usually do not provide usable intelligence for the contents and placement of objects in a room or space.